Re: [XenPPC] [rfc][linux] xencomm encapsulates a NULL pointer

[Hollis Blanchard <hollisb@xxxxxxxxxx>]

On Wed, 2006-07-12 at 00:18 -0400, Jimi Xenidis wrote:
> sometimes the xencomm logic is passed a NULL pointer which is a valid  
> value.
>
> Since the check for NULL is done with:
>    #define guest_handle_is_null(hnd)        ((hnd).p == NULL)
> 
> I figure we should still initialize but mark all array entires as  
> invalid.
> The follow patch should do the trick.
> comments?
> -JX
> 
> diff -r 72c8bc5d88f4 drivers/xen/core/xencomm.c
> --- a/drivers/xen/core/xencomm.c      Thu Jun 29 13:04:30 2006 -0400
> +++ b/drivers/xen/core/xencomm.c      Wed Jul 12 00:15:56 2006 -0400
> @@ -60,7 +60,7 @@ static int __xencomm_init(struct xencomm
>       int i = 0;
>       /* record the physical pages used */
> -     while ((recorded < bytes) && (i < desc->nr_addrs)) {
> +     while (buffer && (recorded < bytes) && (i < desc->nr_addrs)) {
>               unsigned long vaddr = (unsigned long)buffer + recorded;
>               unsigned long paddr;
>               int offset;

Shouldn't 'bytes' be 0 in the case of a NULL pointer? In that case, I
think this code should already work?

-- 
Hollis Blanchard
IBM Linux Technology Center


_______________________________________________
Xen-ppc-devel mailing list
Xen-ppc-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ppc-devel